OpenComputer
github.com
- Free plan
- Windows
- macOS
- Linux
- API
What OpenComputer is
OpenComputer is a free, open-source desktop agent that runs locally, understands the screen, and operates applications through mouse and keyboard interactions. It can click, drag, scroll, enter text, use shortcuts, open URLs, and wait for interface changes. You can connect a local vision-language model through vLLM or use a hosted API, provided the endpoint accepts image inputs, function tools, and the model’s tool-call format. Ask first is the default permission mode and requires approval for mouse, keyboard, and URL actions; Automatic mode acts without individual prompts. Build targets include Linux AppImage and deb packages, Windows NSIS, and macOS DMG. Linux X11 is best supported, while Wayland, Windows, and macOS are experimental. The agent currently works only on the primary display. Screenshots, prompts, and tool results can remain on the machine. API keys use per-user file permissions but are not stored in the operating-system keychain. The project describes itself as early alpha.
Who it suits
It suits users who want local desktop automation with screen understanding and approval controls. It is aimed at people comfortable supervising an early-alpha tool and configuring a compatible model endpoint.
Good to know, and what to watch
- Open-source and free
- Runs locally on Linux, Windows, and macOS
- Supports mouse, keyboard, URL, and scrolling actions
- Ask first approval mode is enabled by default
- Screenshots and prompts can remain on the machine
- Only the primary display is supported
- Wayland, Windows, and macOS are experimental
- API keys are not stored in the OS keychain
- Automatic mode uses the user's desktop privileges
The verdict
OpenComputer provides local, screen-aware desktop control with configurable approvals and model choices. Its early-alpha status, experimental platform support, single-display limit, and security caveats call for supervision.
What the maker publishes
- Action review
- Ask first is the default permission mode and requires approval for mouse, keyboard, and URL actions; Automatic mode runs actions without individual prompts. source
- API compatibility
- The configured endpoint must support image inputs, function tools, and the model’s tool-call format. source
- Approvals
- Ask first is the default permission mode and requires approval for mouse, keyboard, and URL actions. source
- Credential storage
- API keys are stored in the per-user application-data directory with user-only file permissions, but are not yet backed by the operating-system keychain. source
- Desktop controls
- It supports clicking, dragging, scrolling, text entry, keyboard shortcuts, opening URLs, and waiting for UI changes. source
- Display limit
- The current agent operates only the primary display, and multi-monitor automation is on the roadmap. source
- Installer formats
- Build targets are Linux AppImage and deb, Windows NSIS, and macOS DMG. source
- Integrations
- The current model boundary uses an OpenAI-compatible Chat Completions API; dedicated Discord, search, and MCP integrations are listed as future work. source
- License
- OpenComputer is licensed under the MIT License, while third-party dependencies and model weights retain their own licenses. source
- Maker
- The linked GitHub profile identifies the maker as Andrew Krikorian and lists California. source
- Model compatibility
- The model endpoint must support image inputs, function tools, and the model's tool-call format. source
- Model options
- It can use a local vision-language model through vLLM or a compatible hosted API. source
- Privacy and security
- The README says screenshots, prompts, and tool results can stay on the user's machine, and that credentials, screenshots, model calls, and automation remain in Electron's main process. source
- Project maturity
- The project describes itself as early alpha and warns that desktop actions should be supervised, especially around sensitive or irreversible workflows. source
- Purpose
- OpenComputer is an open-source, local-first desktop agent that understands the screen and operates applications through mouse and keyboard interactions. source
- Security design
- The Electron renderer is sandboxed with Node.js integration disabled, while privileged model and desktop automation behavior runs in the main process. source
- Security limitations
- The project says its controls are defense in depth rather than a complete sandbox, and Automatic mode lets the model act with the user's desktop privileges. source
- Security reporting
- The security policy asks users to report vulnerabilities privately and says reports should receive an initial response within seven days, though response times may vary because it is a volunteer project. source
- Supported systems
- Build targets are Linux AppImage/deb, Windows NSIS, and macOS DMG; Linux X11 is best supported, while Wayland, Windows, and macOS are experimental. source
- What it does
- OpenComputer is an open-source desktop agent that understands the screen and can operate applications with mouse and keyboard interactions. source
Checked against the maker's pages on 7 Oct 2026.